Many studies have added to our understanding of key similarities and differences between natural non-progressive and progressive HIV and SIV infections. These studies have highlighted potential explanations for the lack of disease progression in SIV-infected natural hosts. Importantly, it is clear that immunological control, in terms of a virus-specific T cell and B cell response, does not account for the lack of disease progression. Indeed, SIV-infected natural hosts maintain high plasma viral loads (based on amounts of viral RNA) (Goldstein et al., 2006; Goldstein et al., 2000; Pandrea et al., 2006a; Pandrea et al., 2006b; Silvestri et al., 2003) and they do not exhibit superior cellular control of viremia compared to HIV-infected humans or SIV-infected rhesus macaques (RM)(Dunham et al., 2006).

The lentiviruses that infect natural hosts are themselves clearly pathogenic in certain contexts; indeed, SIV infection of African green monkeys (AGM) and Sooty mangabeys (SM) has been associated with more rapid death of infected cells in vivo (Gordon et al., 2008; Klatt et al., 2008; Pandrea et al., 2008a). Moreover, SIVagm, which naturally infects AGM, can be used to infect pigtail macaques who subsequently develop simian AIDS (Goldstein et al., 2005; Hirsch et al., 1995) and isolates of SIVsmm, which naturally infects SM, also cause progressive infection in RM (Fultz et al., 1989; Hirsch et al., 1997; Li et al., 1992; Watson et al., 1997). Therefore, the lack of disease progression in natural hosts is unlikely to be attributable to infection with nonpathogenic viruses. However, analysis of certain viral accessory proteins, such as nef, has led to the suggestion that differential down-regulation of certain key surface proteins on infected cells or differential recruitment of immune activation adapter proteins or binding to host restriction factors by SIV-specific nef variants could influence disease progression (Arhel and Kirchhoff, 2009).

A fundamental difference between progressive SIV or HIV infection and non-progressive SIV infection is the absence of immune activation during the chronic phase of infection in natural hosts (Chakrabarti et al., 2000; Pandrea et al., 2006a; Pandrea et al., 2003; Pandrea et al., 2007b; Silvestri et al., 2003; Sumpter et al., 2007). The importance of this observation rests on the findings that chronic systemic immune activation is associated with disease progression in HIV-infected individuals and is a better predictor of outcome than plasma viral load (Deeks et al., 2004; Giorgi et al., 1999; Rodriguez et al., 2006). The lack of immune activation during the chronic phase of natural SIV infection is therefore thought to be critical for the non-progressive nature of the infection (Silvestri, 2005; Silvestri et al., 2007) but might be somewhat surprising given that natural hosts mount a robust immune response, including production of type I interferons and increased frequencies of activated T cells, during the acute phase of SIV infection (Bosinger et al., 2009; Jacquelin et al., 2009). Moreover, in vitro production of type one interferons by SIVagm-stimulated AGM plasmacytoid dendritic cells are enhanced compared to SIVagm stimulation of RM plasmacytoid dendritic cells, while such innate immune responses to SIVmac are greater in RM compared to AGM (Jacquelin et al., 2009). However, in contrast to SIV or HIV infection of RM and humans, respectively, this immune activation resolves towards the end of the acute phase of infection in vivo in the natural hosts, and may be in part due to differential up-regulation of the membrane receptor programmed death 1 in lymph nodes in of acutely-infected natural hosts (Estes et al., 2008) and TNF related apoptosis-inducing ligand after in vitro infections (Kim et al., 2007).

Several studies have tried to elucidate the mechanisms underlying the differential immune activation in progressive and natural HIV and SIV infections. It has recently been shown that one cause of immune activation during the chronic phase of progressive infection is microbial translocation (Baroncelli et al., 2009; Brenchley et al., 2006b; Ciccone et al.; Funderburg et al.), which refers to the translocation of microbial products from the lumen of the intestinal tract into the systemic circulation. Microbial products can directly stimulate the immune system via interactions with pattern recognition receptors such as Toll-like receptors. Indeed, elevated amounts of soluble CD14, a protein produced by LPS-stimulated macrophages and monocytes, in the plasma of chronically HIV-infected individuals suggest that macrophages and monocytes are directly stimulated by LPS in vivo (Brenchley et al., 2006b). Moreover, elevated amounts of translocated microbial products are associated with immune activation of both the adaptive and innate arms of the immune system during the chronic phase of progressive HIV or SIV infection (Brenchley et al., 2006b). In stark contrast, microbial products are present at only very low amounts in chronically SIV-infected AGM (Pandrea et al., 2007b) and SM (Brenchley et al., 2006b) even in the presence of high viral loads. Thus, the lack of microbial translocation may in part account for the lack of systemic immune activation.

The mechanisms underlying microbial translocation during the chronic phase of progressive HIV or SIV infection and the lack thereof in non-progressive infection remain unclear but several plausible explanations exist. Massive damage to many components of the mucosal barrier of the GI tract, both immunologic and structural, occurs in the acute phase of the infection and persists through to the chronic phase in progressive HIV and SIV infection. Enteropathy with increased intestinal permeability has been routinely observed in chronically HIV-infected individuals (Bjarnason et al., 1996; Kapembwa et al., 1991; Kotler et al., 1984) and considerable enterocyte apoptosis is observed during the acute phase of SIV infection in RM (Li et al., 2008). In addition there is massive infection and loss of CD4⁺ T cells, particularly from the lamina propria, in acute infection of RM and humans(Brenchley et al., 2004; Guadalupe et al., 2003; Mehandru et al., 2004; Veazey et al., 1998). Thus, it has been proposed that the structural and immunologic damage to the gut in progressive HIV and SIV infection allows for microbial translocation to increase and contribute to persistent immune activation even though the high viral loads of the acute phase wane as the infected individual enters the chronic phase of the infection (Brenchley et al., 2006a; Pandrea et al., 2008b). Indeed, the consequences of such microbial translocation might offset any normal physiological attempts at immunological resolution of the acute immune activation that is likely driven in large part by high viral loads.

One might expect, therefore, that SIV-associated damage to the GI tract is minimal in natural hosts compared to hosts with progressive disease. Yet, recent studies actually demonstrated a substantial reduction in the frequency of CD4⁺ T cells at the mucosal surfaces of acutely SIV-infected SM (Gordon et al., 2007) and AGM (Pandrea et al., 2007b). However, more detailed analyses have revealed both quantitative and qualitative differences in CD4⁺ T cell depletion between natural and non-natural hosts that highlight the importance of target cell availability and depletion in determining disease progression. As mentioned above, several studies have shown that natural hosts differentially regulate the cell surface receptors for SIV. CD4⁺ T cells in all species of natural hosts that have been studied thus far have been shown to express very low amounts of CCR5 whereas CD8⁺ T cells appear to express CCR5 amounts comparable to those in humans and RM (Pandrea et al., 2007a). Whereas such low expression of CCR5 by CD4⁺ T cells has been suggested as a mechanism underlying the lack of vertical transmission in natural hosts (Pandrea et al., 2007a), one might also expect that low CCR5 expression would translate to reduced target cell availability, particularly in the GI tract. However, given the lack of disease progression in natural hosts the biological relevance of the lack of vertical transmission remains unclear and requires further studies of infection in infant natural hosts. Importantly, in addition to low amounts of CCR5, AGM down-regulate CD4 expression as naïve CD4⁺ T cells are activated to enter the memory T cell pool (Beaumier et al., 2009). The functional properties usually associated with memory CD4⁺ T cells are assumed by these CD4-negative T cells, thus ensuring the immunocompetence of AGM. The consequence of CD4 down-regulation in these natural hosts would be a reduction in the availability of target cells for SIV infection. Indeed, uninfected healthy adult AGM have few memory CD4⁺ T cells circulating or in the tissues (Beaumier et al., 2009; Pandrea et al., 2007b) and SM have less overall CD4⁺ T cell depletion upon SIV infection compared to either HIV-infected humans (Brenchley et al., 2004) or SIV-infected RM (Brenchley and Douek, 2008; Picker et al., 2004; Veazey et al., 1998). Thus, whereas depletion of CCR5⁺ CD4⁺ T cells in SIV-infected natural hosts may be important in terms of frequency within that subset (whatever CCR5⁺ CD4⁺ T cells targets exist are rapidly lost), their depletion in terms of absolute numbers, or frequency within the total T cell pool, may actually be so minimal that it is of no physiological or pathogenic consequence (Figure 1). However, natural or experimental SIVsmm-infection of sooty mangabeys with a rare CCR5-CXCR4-CCR8-tropic SIV results in severe depletion of mucosal and peripheral CD4⁺ T cells with no apparent susceptibility to opportunistic infections (Milush et al., 2007). The mechanisms by which these animals preserve immunological function and avoid disease progression remain unclear.

In addition to these quantitative differences, there may be qualitative differences in CD4⁺ T cell depletion that may contribute to the lack of disease progression in natural hosts. Indeed, recent studies have shown that CD4⁺ T cells producing IL-17 (Th17 cells) are preferentially depleted from gastrointestinal tracts of chronically HIV-infected humans (Brenchley et al., 2008) and SIV-infected RM (Cecchinato et al., 2008), but that Th17 cells are present at healthy frequencies in the GI tracts of chronically SIV-infected SM (Brenchley et al., 2008) and AGM (Favre et al., 2009). Th17 cells recruit neutrophils and myeloid cells to effector sites by inducing granulocyte-colony stimulating factor secretion (Ye et al., 2001), and are critical to epithelial cell regeneration in mucosal tissues (Brand et al., 2006) partly through the induction of expression of claudins, components of epithelial tight junctions. Th17 cells also induce epithelial cell production of defensins and mucin, both vital to the integrity of mucosal barrier (Chen et al., 2003; Sugimoto et al., 2008). Thus, preferential loss of Th17 cells from the GI tract during progressive HIV and SIV infections might lead to reduced structural and immunologic integrity of the mucosal barrier with a consequent increase in microbial translocation. Conversely, their preservation in SIV-infected natural hosts may partly explain the lack of microbial translocation observed in such animals. Therapeutic interventions such as IL-23 and retinoic acid that aim to increase Th17 cell frequencies in progressive infection and IL-1 receptor antagonists which aim to decrease Th17 cell frequencies in non-progressive infections could shed further light on the biological significance of Th17 cells in the maintenance of the tight epithelial barrier of the GI tract and microbial translocation. Epithelial damage could, in turn, also lead to decreased production of secretory IgA leading to decreased clearance of translocated microbial products.

While, recent studies have clearly demonstrated increased enteroycte apoptosis during the acute phase of SIV infection (Li et al., 2008), the mechanisms underlying this enterocyte apoptosis are unclear. Potential mechanisms underlying enterocyte apoptosis could involve lentiviral envelope proteins binding to oligosaccharide moieties present on a G coupled receptor GPR15 (Maresca et al., 2003) and recent studies have implicated TNF in the HIV-mediated enterocyte damage (Nazli et al.).

Thus, we propose that a critical factor that differentiates nonprogressive from progressive HIV or SIV infection is the maintenance of T cell immune competence during and after the acute phase of disease in the face of a virus that infects and kills CD4⁺ T cells. Hence natural hosts may have adopted mechanisms by which target cell availability for SIV is reduced while immunological function is maintained by subsets of T cells that are resistant to infection. Importantly, down regulation of CCR5 by all natural host species studied and down-regulation of CD4 by AGM occurs irrespective of SIV infection. Although the molecular mechanisms by which CD4 in AGM and CCR5 in all natural hosts are differentially expressed compared to other primate species remain unclear, these findings suggest the co-evolution of natural hosts with SIV.

The mitigation of lentivirus infection of CD4⁺ T cells through downregulation of the receptors for the virus raises an interesting paradox that clearly needs some explanation, namely: how are high plasma virus loads maintained in natural hosts if target cell availability has been reduced? In both natural hosts and hosts with progressive infection an equilibrium is set up between the availability of target cells, the depletion of target cells and the production of infectious virions. In progressive disease, this equilibrium is unstable and decays towards a state of loss of immune competence related to the reduced provision of effector memory CD4⁺ T cells that protect from pathogens that define AIDS. In contrast, this equilibrium appears to be maintained for the normal lifespan of natural hosts with no apparent detriment to their health. We propose that such a stable equilibrium may be achieved through the “uncoupling” of virus production from the processes that lead to disease progression. To understand how this might occur we should first consider the maturation of CD4⁺ T cells into their different subsets. Upon stimulation through the TCR, naïve CD4⁺ T cells may mature into different memory T cell subsets as shown in Figure 2 top. Central memory T cells are thought to have stem cell-like qualities in that they may be regarded as a self-renewing source of effector memory and activated T cells (van Grevenynghe et al., 2008). Central memory T cells tend to be localized to organized lymphoid tissues such as lymph nodes and gastrointestinal Peyer’s patches, whereas effector memory T cells tend to be localized to peripheral tissues such as the broncoalveolar fluid and gastrointestinal lamina propria. Both of these T-cell subsets are in a “resting” state and thus their infection with HIV or SIV would result in the production of small numbers of virions. However, activated memory T cells, defined by the expression of Ki67, may produce up to six-fold more virions per cell (Reilly et al., 2007). Thus the target cells responsible for the production of the majority of virions in a chronically-infected individual would be contained within the activated memory T cell subset (Figure 2 center, bottom). Indeed, it has been shown that the availability of activated CD4⁺ T cells in naturally SIV-infected sooty mangabeys dictates the degree of viremia (Klatt et al., 2008) (Figure 2 bottom). When compared to central memory T cells, both effector memory and activated memory T cells have a far greater tendency to activation-induced cell death; thus, theoretically, their death from viral infection should not have a great impact on their steady state numbers while enabling the production of great quantities of virions. By contrast, the loss of T cells from central memory pool would, over time, severely limit the numbers of effector memory and activated memory T cells available to control pathogens in the tissues. Thus it is the loss of central memory CD4⁺ T cells which introduces instability into the equilibrium between the availability of target cells, the depletion of target cells and the production of infectious virions, and progression to AIDS occurs because of the eventual failure of the provision of effector memory CD4⁺ T cells to the tissues (Okoye et al., 2007).

Thus one might consider HIV and SIV infection as having two separable components: disease progression which is dependent on infection and depletion of central memory CD4⁺ T cells, and virus production which is dependent on infection of activated memory T cells. A survival mechanism could therefore have evolved in natural hosts to uncouple disease progression from virus production by reducing infection and depletion of central memory CD4⁺ T cells while maintaining viral loads through infection of activated CD4⁺ T cells (Figure 2 bottom) (Paiardini et al., 2009). The critical aspect of such uncoupling is to alter target cell availability and we have described above how this has been achieved in AGM and SM by two different mechanisms. In progressive disease of RM and humans any activating stimuli generate target cells within the central memory T cell population by increasing CCR5 expression whereas in natural hosts any activating stimuli remove target cells from central memory CD4⁺ T cells by reducing CCR5 in SM and CD4 in AGM, thus preserving this critical T cell subset while maintaining effector cells that produce virus and thus maintain plasma viral loads. However, the amount of virus production by individual subsets of CD4⁺ T cells in progressive and nonprogressive infections is not well understood. Thus, one critical question that must be further investigated is how much virus is produced from given infected CD4⁺ T cell subsets in order to maintain particular viral loads in different hosts.